Bone cancers destroy bone (tumor osteolysis), causing pain, decreased mobility and, ultimately, skeletal fracture. Current treatments for tumor osteolysis are only palliative and not preventive. It is known that tumors destroy bone by inducing the formation and activation of osteoclasts, however, the mechanisms through which bone tumors direct this influence are unknown. A primary obstacle to developing effective treatments for tumor osteolysis is the absence of knowledge defining tumor-secreted cytokines that induce formation and activation of osteoclasts. As emerging methods of cancer therapy are capable of directing cytokine-specific treatments at cytokine-mediated pathological processes, identification of a cytokine secreted by osteolytic bone tumors, which causes formation and stimulation of osteoclasts, would serve as a critical first step in designing new treatments for patients with this devastating bone disease. The hypothesis which will be tested in this proposal is that a tumor-secreted cytokine, macrophage colony-stimulating factor (M-CSF), causes tumor osteolysis. This postulate is based on the following findings: 1) M-CSF is required for osteoclast formation under physiological conditions; 2) M-CSF is produced by tumors which destroy bone and induce the formation and activation of osteoclasts in vivo; 3) M-CSF is not secreted by tumors which do not destroy bone and do not induce the formation or activation of osteoclasts in vivo; and 4) other cytokines (e.g., IL-6, TNF and GM-CSF) that are known to induce formation and activation of osteoclasts are not produced by these tumors. In this proposal, the applicant will initially develop subclones of these tumor cells, which have the opposite M-CSF-secreting capacity as their corresponding parent cells, using retroviral-mediated gene transfer techniques. Subsequently, he will test the influence of these subclones, carrying either antisense or sense M-CSF genes, on osteoclasts in vivo.